noisy t/c signal

[Quix]

Hi,

I'm simply trying to get a t/c signal from a two wire T type (really fine gauge) which is unsheilded and I'm using differential wiring.

I followed the manual's requirement regarding placing a resistors, I've taken them out, put them in, changed the channel, stood on one foot and hopped.

But just touching the t/c end I get huge noise--to the point its recording -3M degrees C in spikes all over the place in my data.

I don't want to put an RC network in to smooth it out because I need the fast time constant from the small wire.

Any suggestions?  I'm hoping I'm just doing something stupid and don't see it.

Thanks,

Shawn 

[johnsold]

Shawn,

While standing on one foot and hopping you also need to rub your tummy and pat your head... with your arms crossed behind your back.

What is happening is that by touching the t/c you are coupling from ambient electric or magnetic fields, probably at the power line frequency, into your temperature measurement system.  The t/c signal is only a few millivolts or less at temperatures where you can touch the t/c.  Depending on the fields in the area, you can easily touch-couple signals on the order of volts.

Please describe what you are measuring and what other things are in the same area, especially things which use a significant amount of power.  How fast do your temperature measurements need to be?  What DAQ device, computer, and software are you using?

Lynn

[Quix]

Hi Lynn,

Thanks What I ended up doing is taking shielded cable then extending the small t/c wire from this about 6"...  This seems to have worked well.  I get about 70ms to 90% of the signal but also don't have the noise.

Also, I found using FFT on the data the signal was a 7.8Hz, very strong.  Something I've never seen because I hace never needed the fast time constant.

Thanks again for the response.

Regards,

Shawn

[johnsold]

Shawn,

The interfering signal is probasbly not at 7.8 Hz.  Most likely you are seeing an alias of the real signal due to low sampling rate.  If you have eliminated the noise, then you really do not need to worry about it.

Is the time constant still a problem? 

Lynn